This study describes service users with Down syndrome (N = 1,199) and a comparative sample with intellectual and developmental disabilities but not Down syndrome (N = 11,182), drawn from National Core Indicator surveys of adult service users in 25 U.S. states. Individuals with Down syndrome were younger than were individuals without Down syndrome. Men with Down syndrome were older than women with Down syndrome, whereas the reverse was true of the individuals without Down syndrome. Most (68%) people with Down syndrome had mild or moderate intellectual disability. The prevalence of vision impairment, hearing impairment, and physical disability increased with age. Adults with Down syndrome were more likely to have Alzheimer's dementia, have a hearing impairment, or be overweight, but they were less likely to have a physical disability than those without Down syndrome. Adults with Down syndrome were less likely to live in institutions or their own home, but they more likely to live in a family member's home. The results of a logistic regression showed that participants were more likely to be reported to be overweight if they had Down syndrome, were female, and were physically inactive, but they were less likely to be reported to be overweight if they were older, had more severe intellectual disability, had cerebral palsy, or were not independently mobile.
Down syndrome is the best known and most prevalent biologically based condition associated with intellectual and developmental disabilities (IDD). It is caused by an unexplained chromosomal anomaly resulting in 47 chromosomes rather than the usual 46. The extra chromosome affects the typical patterns of body and brain development. The Centers for Disease Control and Prevention in the United States (2006) estimated that one in 733 live births is a child with Down syndrome. In addition to genetic testing, clinical indicators of Down syndrome include poor muscle tone and significant physical and cognitive developmental delay. Most individuals with Down syndrome have intellectual disability (ID), with most testing in the mild or moderate range (Batshaw, 2007).
Down syndrome has been the focus of substantial developmental, educational, and medical research. Although Down syndrome is consistently associated with developmental challenges, environmental factors have been demonstrated to be a mediating influence on development. Infant stimulation and early intervention programs have well-established effects on the development of language and speech, motor skills, cognition, and social skills. Continuing special education has confirmed effects on development and learning (Rynders & Horrobin, 1996).
Individuals with Down syndrome have a high incidence of heart defects and a greater susceptibility to respiratory infections and are at significantly higher risk of Alzheimer's disease than the general population (e.g., Zigman et al., 2007). Among a sample of 89 adults with Down syndrome in a region of the United Kingdom, chart reviews revealed that 36% had ophthalmic disorders such as cataracts; 33% had a hearing deficit; 28% had seizures (two thirds with adult onset); 23% had hypothyroidism; 23% had eczema; 20% were classified as obese; 19% had depression; 16% had dementia; and 58% of adults 40 years or older had Alzheimer's disease (Henderson, Lynch, Wilkinson, & Hunter, 2007). Improved medical care has resulted in an extraordinary increase in life expectancy over the past 30 years (Bittles & Glasson, 2004), with life expectancy for children born today with Down syndrome exceeding 60 years. Yang, Rasmussen, and Friedman (2002) found that between 1983 and 1997, the median age at death for people with Down syndrome increased from 25 years to 49 years, or an average of 1.7 years per year. Increased life expectancy means that many people with Down syndrome are now living long enough to develop age-related health conditions, which in turn affect their health care, support needs, and lifestyle.
As the life expectancy of people with Down syndrome increased, more adults with Down syndrome entered the adult community services systems of state developmental disabilities services. Their increased longevity and the heightened attention to their well-being have brought new concerns, such as high rates of overweight and obesity (Hove, 2004; Rimmer, Heller, Wang, & Valerio, 2004; Robertson et al., 2000), a high prevalence and early onset of Alzheimer's disease (Henderson et al., 2007; Holland, Hon, Huppert, Stevens, & Watson, 1998; Prasher, 1995a; Visser et al., 1997), and other indicators of premature aging.
Adults with Down syndrome are more likely than the general population to be overweight, with individuals living with family members being more likely to be overweight than those living in community group homes (Rubin, Rimmer, Chicoine, Braddock, & McGuire, 1998). In a U.K. study, nearly half (48%) of 201 individuals with Down syndrome were obese, and an additional one in four were overweight (Prasher, 1995b). Again, those living with family members were more likely to be overweight than those living in residential settings. An interesting result was that adults with Down syndrome who had total cholesterol levels of 200 mg/dL or more were twice as likely as those with lower levels to develop Alzheimer's disease (Zigman et al., 2007).
Despite the considerable attention to people with Down syndrome in medical and developmental research, less attention has been paid to the social services and supports received by adults with Down syndrome, for several possible reasons. First, although Down syndrome is the most frequently identified biological condition associated with ID, the estimated 400,000 people with Down syndrome in the United States make up only about 14% of the estimated 2.8 million Americans with ID (Larson et al., 2001; March of Dimes, 2010; National Down Syndrome Society, 2010). Their overall low prevalence in most study samples yields a group too small for other than basic descriptive analyses.
Moreover, adults with Down syndrome are relatively more likely to live with family members and less likely to be in out-of-home placements than other people with IDD (Prasher, 1995c). The reasons for this are not clearly established, but when compared with other families with offspring with ID, those with offspring with Down syndrome have reported more child-related rewards and less stress (Hodapp, Ly, Fidler, & Ricci, 2001). In recent years, adult service systems for people with IDD have substantially increased access to in-home family supports (Lakin, Larson, Salmi, & Webster, 2010). This trend, together with increased longevity, suggests that the number of adults with Down syndrome receiving IDD services is likely increasing.
Because of the relatively low prevalence of Down syndrome, research on the status, experiences, and outcomes for adults with the condition requires a large data set with substantial capacity to integrate diagnostic, functional, health, behavioral, service setting, and outcome variables. In this study, we created such a data set by merging individual state data collected as part of the National Core Indicators (NCI) program. In all, data from 25 states were merged to create an overall sample of 12,381 randomly sampled adults receiving services under the auspices of state IDD service agencies, of whom 1,199 were people with Down syndrome.
Given the substantial increase in life expectancy among people with Down syndrome noted previously, it is important to document service user numbers, health conditions, and impairments across the adult life span. Therefore, a fundamental purpose of this study was to analyze these variables by age group among adults with and without Down syndrome. In addition, we considered living arrangements to be important because of (a) the previously reported differences in obesity by living arrangements (Prasher, 1995b; Rubin et al., 1998) and (b) the greater likelihood that adults with Down syndrome live with family members. Therefore, we documented the type of living arrangements used by adults with and without Down syndrome and undertook comparisons by living arrangements of important health indicators such as overweight. Our overall purpose was to provide descriptive data about age and age-related impairments and health conditions (including Alzheimer's disease) and living arrangements among adult service users with Down syndrome and a comparative sample of adults with IDD but not Down syndrome.
The NCI program is an annual quality assurance endeavor to evaluate performance and outcomes of IDD service systems in participating states. The NCI is a collaboration between the National Association of State Directors of Developmental Disabilities Services and the Human Services Research Institute. No NCI data are collected in nonparticipating states.
Data were drawn from 25 of the 26 states that took part in the NCI program in 2006–2007 and 2007–2008. The exception, Maine (N = 436), did not gather data about Down syndrome diagnosis. For the 23 states that participated in both years, we used the more recent 2007–2008 data. For the two states that took part only in 2006–2007 and not in 2007–2008, we used the 2006–2007 data (these two states conducted NCI surveys every second year). The 25 participating states were Alabama, Arkansas, Arizona, California (Orange County Regional Center only; 2006–2007), Connecticut, Delaware, Georgia, Hawaii, Indiana, Kentucky, Louisiana, Missouri, North Carolina, New Jersey, New Mexico, New York, Oklahoma, Pennsylvania, Rhode Island, South Carolina, Texas, Vermont, Washington (2006–2007), West Virginia, and Wyoming. State samples were randomly selected from each state's population of adults (age 18 and older) with IDD receiving institutional or community- or home-based services, or some subset of these (e.g., some states restricted their sample to recipients of home- and community-based services).
Table 1 presents data on the size of NCI samples in each state, the number and proportion of sample members with Down syndrome, and the estimated prevalence of people with Down syndrome in 25 state developmental disabilities service systems as adjusted on the basis of variations among states in sample size and proportion of total service recipients included in the NCI sample. The second column of Table 1 (labeled column A) indicates the size of the NCI sample for each individual state. As shown in column A, most states in the study had about 400 consumer participants (18 of 25 states with 400 or more participants, SD = 90). Because of this, states contributed a fairly consistent proportion of the total sample members in the study, with 17 of the states contributing 3%–4% of the total sample, from a low of 1.1% in West Virginia to a high of 12.9% in Texas (column B). Of course, the states varied considerably in terms of state population. As shown in column C, the individual state samples varied from 2.4 sample members per 100,000 of the total state population in New York to 73.7 sample members per 100,000 in Wyoming (U.S. Census Bureau, 2011).
States, largely related to their overall populations, varied considerably in their total number of service recipients (column D), defined here as the number of people in the state with developmental disabilities receiving intermediate care facilities for people with mental retardation (ICF/MR) and home and community-based services on June 30, 2009 (Lakin et al., 2010). States ranged from 951 service recipients in Delaware to 69,859 in New York, or expressed in another way, the proportion of the total number of service recipients in all participating states (313,169) from each individual state varied from 0.3% in Delaware to 22.3% in New York (column E). Because of this, the relative contribution of the various states to the overall population estimates differed considerably.
Within the total sample, 1,199 individuals were identified as having Down syndrome, ranging from 11 in Delaware to 133 in Pennsylvania (column F). The primary factor in the number of individuals with Down syndrome among each state sample was the actual size of that sample. For example, Delaware and West Virginia had the smallest numbers of sample members with Down syndrome (11 and 20, respectively) and the smallest overall samples (234 and 137, respectively), whereas Pennsylvania and Texas had the highest numbers of sample members with Down syndrome (133 and 107, respectively) and the largest numbers of total sample members (1,224 and 1,594, respectively). However, the difference in the proportions of people with Down syndrome within the individual state samples ranged from 14.6% in Vermont and West Virginia and 14.4% in Orange County, California, to 4.7% in Delaware, 5.4% in South Carolina and 6.7% in Texas (column G).
The question that arises from such variation is whether it indicates a problem in the representativeness of the total sample, or whether it more likely derives from the normal variations in sampling when drawing relatively small samples across states. In response we created a weight (shown in column H) to adjust the number of people identified with Down syndrome within the individual states by dividing each state's percentage of total service recipients (column E) by its percentage of the total NCI sample (column B). The actual number of participants with Down syndrome (column F) was then multiplied by this weight (column H) to compute the weight-adjusted number of participants with Down syndrome in each state (column I). Applying this weight increased the estimated number of people with Down syndrome from 1,199 to 1,286 (an increase of 7.3%). It increased the estimated prevalence of people with Down syndrome within state developmental disabilities services systems from 9.7% to 10.4%, which is within the 99% confidence interval ([9.0, 10.4]) for the 9.7% raw prevalence figure. Given the relatively small effect the adjustments made on the total estimated prevalence, we determined that it is most likely that interstate variations are due to sampling effects and that the 25-state sample needed no state-related weights. Therefore, we conducted all remaining analyses using the unweighted raw data.
Data were collected using the NCI Consumer Survey 2006–2007 version (two states) and 2007–2008 version (24 states). The survey is used to gather data about participating individual service users and the services they receive. We drew only on data from the NCI Background section, which was identical in both NCI versions. The NCI Background section requests data on the service user's personal characteristics, functioning, diagnoses, health, problem behavior, living arrangements, and services. These data are obtained from individual records, setting administrators or case managers, and direct support providers.
To ensure that all interviewers receive consistent training, the NCI Consumer Survey protocol is supported by a training program for interviewers, including training manuals, presentation slides, training videos, scripts for scheduling interviews, lists of frequently asked questions, and picture response formats. The training includes question-by-question review of the survey tool.
Multiple tests yielded an interrater agreement of 92%–93%, and a single examination of test–retest reliability resulted in 80% agreement (Smith & Ashbaugh, 2001).
The total NCI sample consisted of 13,169 individual users of adult IDD services from 25 states, but 788 were missing data on Down syndrome diagnosis, leaving 12,381, of whom 1,199, or 9.7% (99% CIs [9.0, 10.4]) were reported to have Down syndrome. The overall sample of 12,381 was made up of 6,855 (55.5%) men and 5,491 (44.5%) women (gender missing for 35 participants), with an average age of 42.7 years (range = 18–100). More detailed information about age, gender, race, level of ID, health, and service use is presented in the Results section.
Where possible, we present effect sizes with the results of statistical analyses. For F tests, we report partial eta-squared (ηP2) and for 2 × 2 chi-square tests, we report the odds ratio (OR).
Sample members with Down syndrome were significantly younger (M = 40.59 years, SD = 11.85, range = 18–82) than those without Down syndrome (M = 42.88, SD = 14.57, range = 18–100), F(1,12,340) = 27.75, p < .001, ηP2 = .002. Age group and Down syndrome were significantly associated, χ2(4, N = 12,342) = 86.05, p < .001. There were fewer individuals in the older age groups for both people with and people without Down syndrome, but this trend was notably more pronounced among sample members with Down syndrome. Adults 60 years and older made up 4.7% of the sample with Down syndrome and 13.5% of the sample without Down syndrome. By comparison, adults 60 years and older constituted 25.6% of the 2008 U.S. population age 18 years and older (U.S. Census Bureau, 2010). Individuals with Down syndrome in the 60 and older age group were also younger, with a mean age of 64.0 compared with a mean age of 67.4 for those without Down syndrome, F(1,1,565) = 12.80, p < .001, ηP2 = .008. Table 2 shows the number and percentage of participants with and without Down syndrome in each age group.
The relative decrease in the percentage of sample members with Down syndrome in older age groups is also shown in Table 2. Although people with Down syndrome made up 9.7% of the overall sample, they represented just 3.6% of those 60 years and older.
Of the 12,346 sample members (99.7% of the total sample) for whom gender was reported, 6,855 (55.5%) were men and 5,491 (44.5%) were women. The sample included 654 men with Down syndrome (9.5% of all men). Men made up 54.7% of all sample members with Down syndrome. The 542 women with Down syndrome in the sample represented 9.9% of all women, and 45.3% of all sample members with Down syndrome. Although gender and Down syndrome status had no general association, the pattern of gender distribution associated with age did show an important difference. This finding was consistent with shorter life expectancy in men without Down syndrome relative to women without Down syndrome, but a longer life expectancy in men with Down syndrome relative to women with Down syndrome. A two-way analysis of variance (ANOVA) of age by gender and Down syndrome status showed that sample members with Down syndrome were significantly younger than those without Down syndrome, F(1, 12306) = 30.79, p < .001, ηP2 = .002. There was no gender main effect (p > .8), but that there was a significant Down Syndrome × Gender Interaction, F (1, 12306) = 6.25, p = .012, ηP2 = .001. For participants without Down syndrome, men's average age (42.4 years) was lower than women's (43.6), whereas for those with Down syndrome, men's average age (41.0) was higher than women's (40.0).
If these differences in average age of male and female sample members arose because of differences in life expectancy, then, other things being equal, we would expect a change in gender mix in the oldest age group relative to the younger age groups that differs in direction by Down syndrome status (see Table 2).
Table 2 shows that for adults without Down syndrome in the 18–29 age group, 59% were male and 41% were female, but in the oldest age group, 53% were male and 47% were female. By contrast, among people with Down syndrome, the gender distribution shifted from approximately 53% male and 47% female in the 18–29 age group to 70% male and 30% female in the 60 years and older age group. These findings suggest that the lower average age of women with Down syndrome relative to that of men with Down syndrome may be the result, at least in part, of gender differences in life expectancy.
Level of ID
Table 3 shows the number and percentage of participants with and without Down syndrome at each level of ID. A small group of participants (4.1% of the total) with no ID label was combined with the group with mild ID. Most (68.3%) people with Down syndrome had mild or moderate ID. There was a difference in the percentages of people with and without Down syndrome by level of disability, with a smaller percentage of people with Down syndrome in the mild and profound levels and a larger percentage with Down syndrome in the moderate and severe levels, χ2(3, n = 11,948) = 219.04, p < .001.
Table 4 shows the number and percentage of participants by Down syndrome status and living arrangement at the time of the NCI interview. The relationship between Down syndrome status and living arrangements was significant, χ2(7, N = 12,311) = 180.32, p < .001. The percentage of people with Down syndrome living in institutions (including ICF/MRs) or their own home was smaller than for the group without Down syndrome. The percentage living in the home of a family member was larger for the group with Down syndrome.
Living with family by age group
Given the large percentage of people with Down syndrome who lived with family members, we examined how common this living arrangement was for different age groups. In young adulthood (18–29) through middle age (40–49), a significantly higher proportion of individuals with Down syndrome lived in the family home, ages 18–29, χ2(1, n = 2,765) = 43.03, OR = 2.52; ages 30–39, χ2(1, N = 2,607) = 37.34, OR = 2.13; and ages 40–49, χ2(1, N = 2.975) = 37.16, OR = 2.06; all ps < .001. In the older age groups (50–59 and older), we found no significant difference in the proportion living with family members by Down syndrome status (see Figure 1).
Stability of living arrangements
Overall, more participants with Down syndrome (72.7%) than individuals without Down syndrome (63.3%) had stable living arrangements, χ2(3, n = 11,841) = 43.05, p < .001. Stability was defined as having lived in one's current home for 6 or more years. Figure 2 shows these percentages by age group and Down syndrome status, together with percentages for sample members who had lived in their current home for less than 1 year.
We found a significant age-related trend toward an increasing percentage of those without Down syndrome having stable living arrangements with older age but the opposite for the group with Down syndrome, χ2(4, n = 7,577) = 79.11, p < .001. A reciprocal trend was evident among those with unstable housing (in current home < 1 year), but it did not attain significance, χ2(4, n = 955) = 4.42, p = .35.
Health Conditions and Impairments
We found no overall difference in the proportion of people with vision impairment by Down syndrome status, χ2(1, n = 12,371) = 1.60, p = .21. Reported vision impairment by age group for people with and without Down syndrome revealed a significantly smaller proportion of people with Down syndrome with vision problems in the 18–29 age group, χ2(1, N = 2,504) = 6.57, p < .01, OR = 0.55, but no significant differences in any other age group. Among those both with and without Down syndrome, the trend was mostly toward a higher percentage of vision problems in the older age groups (Table 5).
Individuals with Down syndrome were more likely to have reported hearing impairments (10.3%) than those without Down syndrome (7.0%), χ2(1, n = 12,370) = 18.25, p < .001, OR = 1.54, with significant differences in the following age groups: 30–39, χ2(1, n = 2,606) = 10.65, p = .001, OR = 2.03; 40–49, χ2(1, n = 2,970) = 3.89, p = .049, OR = 1.47; and 50–59, χ2(1, n = 2,424) = 6.18, p = .013, OR = 1.65. For both those with and without Down syndrome, the trend was toward a higher percentage of hearing problems in the older age groups (Table 5).
People with Down syndrome were less likely to have a physical disability (7.2%) than those without Down syndrome (13.0%), χ2(1, n = 12,371) = 32.41, p < .001, OR = 0.52. This was true for all age groups (see Table 5). For both groups, the percentage with physical disability increased from age 18 to age 59 (see Table 5), but for that with Down syndrome, this percentage declined for the 60 and older age group with Down syndrome, whereas it continued to increase for the group without Down syndrome. We found significant differences in the rate of physical disability by Down syndrome status among the 18–29 age group, χ2(1, n = 2,764) = 20.83, p < .001, OR = 0.18, and the 40–49 age group, χ2(1, n = 2,971) = 7.26, p = .007, OR = 0.58.
Overall, 103 (0.8%) sample members were reported to have Alzheimer's disease. Individuals with Down syndrome (3.4%) were more likely to have Alzheimer's disease than those without Down syndrome (0.6%), χ2(1, n = 12,372) = 108.88, p < .001, OR = 6.40. Figure 3 shows a clear age-related increase in the percentage of those with Alzheimer's disease, but it was noticeably more marked for those with Down syndrome. A significantly higher proportion of people with Down syndrome had Alzheimer's disease for all age groups beginning at age 40: 40–49, χ2(1, n = 2,971) = 21.12, OR = 6.95; 50–59, χ2(1, n = 2,425) = 106.35, OR = 15.59; and 60 and older, χ2(1, n = 1,567) = 47.35, OR = 10.14; all ps < .001.
Individuals with Down syndrome (17.5%) were less likely to be physically inactive than those without Down syndrome (23.9%), χ2(1, n = 10,395) = 22.24, p < .001, OR = 0.67. However, the incidence of physical disability among people without Down syndrome was higher. When we excluded individuals who were not mobile from the analysis, physical inactivity among individuals with Down syndrome (15.6%) and those without (17.6%) no longer differed significantly.
Overall, 27.3% of sample members were reported to be overweight (these data were based on caregiver report of the person being underweight, overweight, or having no weight concerns, not on report of measured weight). Individuals with Down syndrome (38.8%) were significantly more likely to be reported to be overweight than those without Down syndrome (25.9%), χ2(1, n = 10,315) = 78.12, p < .001, OR = 1.81. The difference was significant in the 18–29 age group, χ2(1, n = 2,387) = 36.59, OR = 2.32; the 30–39 age group, χ2(1, n = 2,139) = 27.71, OR = 2.03; and the 40–49 age group, χ2(1, n = 2,499) = 14.80, OR = 1.62, all ps < .001.
Overall, a significantly higher proportion of women (33.6%) than men (22.0%) were reported to be overweight, χ2(1, n = 10,593) = 180.63, p < .001, OR = 1.80. Significantly more men with Down syndrome (33.3%) than without Down syndrome (20.8%) were reported to be overweight, χ2(1, n = 5,681) = 46.63, p < .001, OR = 1.90. Likewise, significantly more women with Down syndrome (45.4%) than without (32.4%) were reported to be overweight, χ2(1, n = 4,600) = 32.04, p < .001, OR = 1.73.
As Figure 4 shows, a larger percentage of women with Down syndrome were overweight than of any other group, with men without Down syndrome the least likely to be overweight (20.8%). Figure 4 does not include a data point for women with Down syndrome age 60 and older because there were too few (n = 13) for a reliable estimate.
Overweight and Alzheimer's
Among individuals with Down syndrome age 40 or older, only 3.8% of those with Alzheimer's reported to be overweight compared with 32.6% of those without Alzheimer's, χ2(1, n = 253) = 9.25, p < .002, OR = 0.08. The two groups did not differ significantly by mean age: 57.2 years for individuals with Alzheimer's; 55.4 years for individuals without Alzheimer's.
Overweight and living arrangements
We analyzed the percentage of people with Down syndrome who were reported to be overweight in various living arrangements: specialized institution (includes ICF/MRs), group home, agency-operated apartment, independent home or apartment (own home), family member's home, and foster care or host home (we did not include specialized nursing facility and other residence types in the analysis because of the very small number of nursing facility residents—0.2% of the sample with Down syndrome—and because residence type was unknown for other). For individuals with Down syndrome, we found a significant univariate difference by type of living arrangement, χ2(5, n = 1,013) = 22.77, p < .001, with a smaller percentage reported to be overweight in institutions and a higher percentage reported to be overweight in agency apartments, own home, and family homes (Table 6).
Although the differences between types of living arrangement in the reported prevalence of being overweight appear to be notable, this finding may be confounded by other variables, in that we also found important variations in participant characteristics between residence types. For example, people living in a family home tended to be younger (see Figure 1) and have Down syndrome, but a higher percentage of people with Down syndrome in the younger age groups were reported to be overweight (see Figure 1). Therefore, to examine the relationship between overweight and living arrangements, it was also important to evaluate simultaneously selected personal characteristics. We conducted a logistic regression analysis with overweight status (yes or no) as the dependent variable and the following independent variables: living arrangement (recoded into binary dummy variables for institution, agency apartment, own home, family home, and host or foster home, with group home as the reference category), age (in years), gender (female = 1), severity of ID (mild = 1, profound = 4), cerebral palsy (yes = 1), nonmobile (not independently mobile = 1), Alzheimer's (yes = 1), physically inactive (inactive = 1), and Down syndrome (yes = 1). This analysis yielded a significant model, χ2(13) = 717.80, p < .001, Nagelkerke R2 = .11. The detailed results are shown in Table 7.
As Table 7 shows, having Down syndrome, being female, and being physically inactive were associated with a higher likelihood of being overweight; and being older, having more severe ID, being nonmobile, and having cerebral palsy were associated with a lower likelihood of being overweight. The magnitude of the effect of physical inactivity was particularly notable, with such individuals being more than twice as likely to be reported to be overweight than their physically active counterparts (OR = 2.55). Relative to group home residents (the reference category), no significant differences by living arrangement type were found in the likelihood of being overweight. We found a trend toward adults living in their own home being more likely to be overweight (OR = 1.14, p = .09), but this trend did not attain significance. Given that we found a Nagelkerke R2 of .11 for the overall logistic regression model, it is not surprising that most effects were small.
We repeated the logistic regression analysis for just participants with Down syndrome and found that the only significant predictors were gender (p < .001, OR = 1.65), severity of ID (p < .05, OR = 0.82), and physical inactivity (p < .001, OR = 2.59; Nagelkerke R2 = .10). Among participants with Down syndrome, being male, being physically active, and having more severe ID were all associated with being less likely to be overweight. None of the living arrangements differed significantly from group homes (the reference category) for the Down syndrome sample.
Overall, adults (18 years and older) with Down syndrome made up almost 10% of users of adult IDD services in our U.S. 25-state sample, based on data collected mostly in 2007–2008. This percentage compares with 13.7% of those with ID (both children and adults) registered with a state disability services commission in one Australian state, using data collected between 1953 and 2000 (Glasson et al., 2003). Adults with Down syndrome and other service users had many similarities but also notable differences that we consider here in greater detail. We discuss our findings, focusing on the areas of demographics, living arrangements, and health conditions and impairments.
Adults with Down syndrome were significantly younger on average than other service users. We found a more marked age-related decline among the percentage of people with Down syndrome, who made up 9.7% of the total sample but only 3.6% of the oldest age group (age 60 and older; Table 2). Moreover, in the 60 and older age group, individuals with Down syndrome were significantly younger than those with other IDD. These findings may be associated with shorter life expectancy among people with Down syndrome (Glasson et al., 2003) and with the effect on life expectancy of the much greater incidence of Alzheimer's disease (discussed below). However, because our study was cross-sectional and involved analysis of the current age of living service users, not their age at death, our data did not directly evaluate life expectancy. We cannot rule out factors other than mortality that may have differentially reduced the number of older people with Down syndrome in the sample (e.g., higher rates of admission of people with Down syndrome to health settings outside the IDD service system). Because our data were cross-sectional, the notable advances in life expectancy for people with Down syndrome in recent decades would have had the least effect on sample members born before 1950 (i.e., those in the age 60 and older age group). These advances will, however, likely increase the future number of people with Down syndrome who attain age 60 and older.
Such considerations aside, our data are consistent with previous findings that individuals with Down syndrome have shorter life expectancy than individuals without Down syndrome and that women with Down syndrome have shorter life expectancy than men with Down syndrome (Table 2). This outcome is consistent with the findings of Glasson et al. (2003), who found a median age of survival for people with Down syndrome of 61.1 years for men but only 57.8 years for women.
Additionally, we found that among people without Down syndrome, the percentage of women increased modestly but consistently with age, in accordance with these women having longer life expectancy than men (as in the general population). Here, too, our findings were consistent with those of Glasson and colleagues (2003), who found a median age of survival for people with ID but without Down syndrome of 68.1 years for men and 74.3 years for women.
We found consistent differences in the type of living arrangements by Down syndrome status. A higher percentage of individuals with Down syndrome (43.3%) than without Down syndrome (27.6%) lived with family, but lower percentages of participants with Down syndrome than individuals without Down syndrome lived in institutions (10.3% vs. 17.6%) and in their own home (7.1% vs. 14.6%); see Table 4. As shown in Figure 1, the likelihood of living with family members was strongly negatively related to the person's age. More important, individuals with Down syndrome were significantly more likely to live with family members in the 18–29, 30–39, and 40–49 age groups, clearly suggesting that they tend to remain living with family members longer in adulthood than individuals without Down syndrome.
Most participants in all age groups had stable living arrangements (i.e., had lived in their current home for more than 5 years), but significantly more sample members with Down syndrome experienced such stability, in large part because of the higher percentage who remained living with family members. However, stability varied with age and Down syndrome status. In younger age groups, a higher percentage of people with Down syndrome had stable living arrangements, but in the 60 and older age group, a higher percentage of people without Down syndrome had stable living arrangements (Figure 2).
For adults without Down syndrome, living arrangements are more likely to change in their younger years, at least partly because they tend to move out of the family home sooner, thus reducing the percentage with stable living arrangements in this age cohort. As they age, their living arrangements become more stable, likely because of the subsequent stability of their out-of-family-home living arrangements. The higher percentage of people with Down syndrome living with a family member in young adulthood and middle age is mirrored in the higher percentage of participants with Down syndrome with stable housing in these age groups (Figure 2). Likewise, among those with Down syndrome age 50 years and older, the rise in the percentage with new living arrangements to which they had moved in the past year could well be associated with the inability of their aging parents to continue to offer support within the family home, especially with the onset of age-related health conditions (including Alzheimer's) among older individuals with Down syndrome.
Our findings are consistent with those of Patti, Amble, and Flory (2010), who reported that among service users with ID age 50 or older, individuals with Down syndrome experienced significantly more changes of residence than those without Down syndrome. Patti and colleagues found that this was partly the result of a much higher percentage of nursing home placements (mostly dementia related) among the Down syndrome group. Nursing home placement was not as evident in our data, likely because many of these health facilities are not part of state IDD systems, so individuals with ID living there would have been excluded from our sample.
Health Conditions and Impairments
For people with and without Down syndrome, we found an age-related trend toward additional impairments in older age, with the prevalence of vision and hearing impairments and physical disability mostly increasing in the older age groups. Vision impairment did not differ by Down syndrome status. However, consistent with previous research (Shott, Joseph, & Heithaus, 2001), hearing impairment was more prevalent among people with Down syndrome (Table 5). By contrast, physical disability was more prevalent among those without Down syndrome (Table 5).
As expected, Alzheimer's disease was markedly more prevalent among people with Down syndrome (OR = 6.40), with a sharp increase in prevalence from around age 50 (OR = 15.59 in the 50–59 age group). Our data involved reports of an existing diagnosis of Alzheimer's disease; participants were not assessed for Alzheimer's as part of the NCI protocol. Therefore, relative to studies that directly assess participants' dementia status, our data likely identified fewer people as having Alzheimer's. In addition, our sample was drawn from current users of IDD services and did not include former service users who, because of health conditions such as Alzheimer's disease, had transferred to health facilities outside state IDD systems (see Patti et al., 2010).
Our findings on the prevalence of Alzheimer's disease in younger adults with Down syndrome (Figure 3) are similar to those of other studies of individuals with Down syndrome (Holland et al., 1998; Prasher, 1995a; Visser et al., 1997). For individuals ages 30–39, figures ranging from 0% to 3.4% have been reported (Holland et al., 1998; Visser et al., 1997), consistent with our finding of 0.7%. However, in older age groups with Down syndrome, our data showed a lower prevalence than other studies. For the 40–49 age group, other researchers reported a prevalence of 9%–11%, and for the 50–59 age group, 36%–66% (Prasher, 1995a; Visser et al., 1997), whereas we found prevalences of 2.3% and 9.1%, respectively, for these age groups. Our highest prevalence finding was 16.1% in the 60 and older age group. Despite these differences, our findings confirm that (a) Alzheimer's disease is a major health concern for adults age 40 and older with Down syndrome, but not for adults in this age group with ID without Down syndrome; (b) the prevalence of Alzheimer's among adults with Down syndrome rises sharply with increasing age; and (c) despite the higher prevalence of Alzheimer's disease, many older adults with Down syndrome do not experience this condition. Neither of the two oldest sample members with Down syndrome (a man age 82 and a woman age 80) was reported to have Alzheimer's disease. Together with our finding that with increasing age, a substantially smaller percentage of adults lived with family, the notable incidence of Alzheimer's among middle-aged and older adults with Down syndrome clearly suggests that community services need to be capable of responding with appropriate support to allow service users with Alzheimer's to age in place (Patti et al., 2010).
Our weight data were based on respondent judgment, not on objective data on body mass index (BMI). Even so, as did Robertson and colleagues (2000), we found that significantly more people with Down syndrome (38.8%) were reported to be overweight than others with ID (25.9%). Using BMI, Rimmer and colleagues (2004) found that 69% of their middle-aged U.S. participants with Down syndrome (M = 39.4 years) were obese, a prevalence that is notably higher than our results. Rimmer and colleagues' sample was specifically selected as being sedentary and had a higher representation of women (56%), who were more likely to live with family members (57% vs. 43% in our sample; see Heller, Hsieh, & Rimmer, 2004), all factors that have been associated with being overweight (see Emerson, 2005).
Our finding that more women with Down syndrome (45.4%) than men (33.3%) with Down syndrome were reported to be overweight was consistent with Hove's (2004) Norwegian BMI data and with Emerson's (2005) and Robertson and colleagues' (2000) data. Robertson and colleagues used U.K. BMI data and found that that more women were overweight in general samples of adults with ID. Moreover, they found female gender and having Down syndrome to be significant independent multivariate predictors of obesity. However, our results differ from those of Prasher (1995b), who reported no gender differences, with 48% of U.K. men and 47% of U.K. women with Down syndrome being obese on the basis of BMI criteria. More important, under multivariate analysis that controlled for age, severity of ID, cerebral palsy, living arrangement type, mobility, Alzheimer's, and physical activity level, we found that being female (OR = 1.79) and having Down syndrome (OR = 1.85) were both significant independent predictors of being reported to be overweight (Table 7).
Among the modifiable factors that we analyzed using logistic regression, physical inactivity was most strongly associated with overweight both for the overall sample (OR = 2.55) and for individuals with Down syndrome (OR = 2.59; note that we had no data on diet). Supporting people with Down syndrome to become more physically active likely represents an important contribution to overall health and may also assist with overweight.
Weight loss and Alzheimer's
Prasher, Metseagharun, and Haque (2004) reported a pattern of progressive weight loss among adults in the 41–60 age range with Down syndrome and Alzheimer's disease that averaged 21% over a 4-year period. Our findings were consistent with Prasher and colleagues' results in that almost none (3.8%) of the individuals with Down syndrome and Alzheimer's were reported to be overweight, compared with 32.6% of those of similar age with Down syndrome but not Alzheimer's. The higher prevalence of Alzheimer's may be one factor contributing to the lower percentage of people reported to be overweight in the older age groups of people with Down syndrome (Figure 4; see also Rubin et al., 1998; Prasher, 1995b).
Overweight and living arrangements
We found significant univariate differences between types of living arrangement in the percentage of people with Down syndrome reported to be overweight, with fewer overweight individuals in institutions and more in family homes, own home, and agency apartments. However, logistic regression analysis showed that with age, gender, severity of ID, Down syndrome status, cerebral palsy, mobility, physical activity, and Alzheimer's all statistically controlled, compared with group home residents, individuals living in other types of setting did not differ significantly in their likelihood of being reported as overweight.
These findings differ from those of BMI-based studies by Rubin and colleagues (1998) and Prasher (1995b) that showed that people living in settings with 24-hour staffing, such as institutions and group homes, are less likely to be overweight than those living with family. However, both studies reported simple univariate comparisons without controlling for personal characteristics. We, too, found similar univariate differences, but under logistic regression these differences by living arrangement were no longer evident, suggesting that personal characteristics such as level of disability and physical inactivity are most important. However, our data were based on respondent judgment, which may not have been as sensitive to differences as BMI. Clearly, the influence of living arrangements on overweight is an important issue that requires future researchers to use both the highest quality data and carefully controlled multivariate analyses.
Once individuals who were not mobile were removed from the sample, reported physical inactivity did not differ between the groups with Down syndrome (15.6%) and without Down syndrome (17.6%). This outcome mirrors Robertson and colleagues' (2000) findings.
Strengths of our study include the large sample size, which allowed for more detailed analyses of associations with Down syndrome status, age, other personal characteristics, and different living arrangements. Furthermore, our sample included (a) random sampling of service users in participating states and (b) data from 25 states. Our examination of the state NCI sample sizes relative to the overall size of each states' developmental disability service system showed that even though there was a degree of proportional over- or underrepresentation of individual states, it made little difference overall to our estimate of the prevalence of Down syndrome in the NCI sample.
One limitation of our study is that we relied on reported data rather than directly assessing each sample member. These reports came from well-informed sources with access to agency records and the individuals' files and included family members, case managers, and administrators. However, some data items were based on informant judgment (e.g., overweight) or records of existing diagnoses (e.g., Alzheimer's) that may not be as accurate as direct, objective current assessment. Where appropriate, we have tried to point out these limitations when discussing our findings. A second notable limitation is the cross-sectional nature of our data. This factor is particularly important when discussing age-related trends because the various age groups represent different cohorts who have had considerably different social, developmental, and service-related experiences in their lives.
Overall, our data have shown that adults with Down syndrome represent about 10% of adult users of IDD services. They are younger on average than other service users and are substantially underrepresented among people 60 years and older, likely because of their shorter life expectancy. Men with Down syndrome appear to live longer than women, but we provided only indirect evidence of this.
As with other individuals with ID, older adults with Down syndrome tend to experience additional impairments as they age. In line with previous research, hearing impairment and Alzheimer's disease were more prevalent among adults with Down syndrome. Individuals with Alzheimer's disease and Down syndrome were less likely to be overweight, consistent with research that has shown notable weight loss as Alzheimer's progresses (Prasher et al., 2004).
Adults with Down syndrome were significantly more likely to be reported as overweight, especially women. A significantly higher percentage of individuals with Down syndrome lived with family members, and fewer lived in institutions or their own homes. Under multivariate analysis, compared with group home residents, we found no significant differences in the chances of being overweight by living arrangements. Finding ways to support people with Down syndrome to avoid becoming overweight represents a significant public health challenge. The finding that adults with IDD who were physically inactive were 155% more likely to be overweight than those who were physically active suggests a great starting point for intervention.
The Human Services Research Institute (HSRI) employs several of the authors. HSRI coordinates the National Core Indicators project and receives a fee from participating states. The authors alone are responsible for the content and writing of the paper.
Preparation of this article was supported by Grant #H133G080029 from the National Institute on Disability and Rehabilitation Research, U.S. Department of Education. Federal funds for this 3-year project total $599,998 (99.5% of the total program costs, with 0% funded by nongovernmental sources).
Editor-in-Charge: Steven J. Taylor
Roger J. Stancliffe (e-mail: firstname.lastname@example.org), Faculty of Health Sciences, University of Sydney, Lidcombe, New South Wales 1825, Australia; K. Charlie Lakin and Sheryl A. Larson, University of Minnesota, Research and Training Center on Community Living; Joshua Engler, Sarah Taub, Jon Fortune, and Julie Bershadsky, Human Services Research Institute.